People love to shit on the idea of using flat-files (plain text files) for things like indexing, but it’s them that are full of shit. Flat files are often the best approach to indexing knowledge because they can be consumed by anything — no really, anything. You can’t say that of an SQLite DB files and the like, hell, you can’t even say that about XML or CSV files.
Flat files also do not require any cognitive overhead or special skills to use and maintain. Anyone who can read and write can make sense of them and use them immediately. Those with minimal scripting skills can iterate over every line in their own way, even import them into spreadsheets or their favorite database engine.
Flat files fulfill the well-established “progressive design” principle. Use the least amount of technology possible and progressively add support for more technology as it is detected, but never assume you have it. Assuming everyone has SQLite, for example, is an example of violating that principle. When it comes to universal, sustainable knowledge storage, sharing, and management we most definitely want the least possible technology requirements.